The effectiveness of a community-based program for reducing the incidence of falls in the elderly: a randomized trial.
ABSTRACT To test whether Stepping On, a multifaceted community-based program using a small-group learning environment, is effective in reducing falls in at-risk people living at home.
A randomized trial with subjects followed for 14 months.
The interventions were conducted in community venues, with a follow-up home visit.
Three hundred ten community residents aged 70 and older who had had a fall in the previous 12 months or were concerned about falling.
The Stepping On program aims to improve fall self-efficacy, encourage behavioral change, and reduce falls. Key aspects of the program are improving lower-limb balance and strength, improving home and community environmental and behavioral safety, encouraging regular visual screening, making adaptations to low vision, and encouraging medication review. Two-hour sessions were conducted weekly for 7 weeks, with a follow-up occupational therapy home visit.
The primary outcome measure was falls, ascertained using a monthly calendar mailed by each participant.
The intervention group experienced a 31% reduction in falls (relative risk (RR)=0.69, 95% confidence interval (CI)=0.50-0.96; P=.025). This was a clinically meaningful result demonstrating that the Stepping On program was effective for community-residing elderly people. Secondary analysis of subgroups showed that it was particularly effective for men (n=80; RR=0.32, 95% CI=0.17-0.59).
The results of this study renew attention to the idea that cognitive-behavioral learning in a small-group environment can reduce falls. Stepping On offers a successful fall-prevention option.
- SourceAvailable from: Catherine Mary Dean[Show abstract] [Hide abstract]
ABSTRACT: Home exercise can prevent falls in the general older community but its impact in people recently discharged from hospital is not known. The study aimed to investigate the effects of a home-based exercise program on falls and mobility among people recently discharged from hospital.PLoS ONE 09/2014; 9(9):e104412. · 3.53 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: In Sweden, a new type of program called ”Active lifestyle all your life” has been developed in primary care. It is a multidisciplinary program aimed at the decrease of accidental falls and at promotion of the health benefits of staying engaged in valued occupations. As one part of the evaluation of this program, this study aimed to explore and understand if, and how, participating in the program influenced each individual's everyday life narratives. Data was collected using a narrative approach with repeated interviews focusing on participation and engagement in valued occupations in the context of accidental fall. Thematic analysis showed that increased awareness related to new ways of thinking and acting among the informants was the over aching theme concerning how the program influenced their everyday life. This increased awareness was experienced as going beyond the cognitive understanding leading to a behavior change and new habits.Physical & Occupational Therapy in Geriatrics 12/2013; 31(4).
Lindy Clemson, BAppSc(OT), MAppSc(OT), P/7D,*
Robert G. Cumming, MBBS, MPH, P/2D,f”
Hal Kendig, MPI, P/2D, FASSA,5 Megan Swann, BAppSc(OT), 7 Robert Heard, BA(Hons), PhD, *
and Kirsty Taylor, BA(Psyc/2)’
OBJECTIVES: To test whether Stepping On, a multifacet-
ed community-based program using a small-group learning
environment, is effective in reducing falls in at-risk people
living at home.
DESIGN: A randomized trial with subjects followed for 14
SETTING: The interventions were conducted in commu-
nity venues, with a follow—up home visit.
PARTICIPANTS: Three hundred ten community residents
aged 70 and older who had had a fall in the previous 12
months or were concerned about falling.
INTERVENTION: The Stepping On program aims to im-
prove fall self-efﬁcacy, encourage behavioral change, and
reduce falls. Key aspects of the program are improving
|ower—limb balance and strength, improving home and
community environmental and behavioral safety, encour-
aging regular visual screening, making adaptations to low
vision, and encouraging medication review. Two-hour ses-
sions were conducted weekly for 7 weeks, with a follow-up
occupational therapy home visit.
RESULTS: The intervention group experienced a 31% re-
duction in falls (relative risk (RR) = 0.69, 95% conﬁdence
interval (CI) = 0.50-0.96; P =.025). This was a clinically
meaningful result demonstrating that the Stepping On pro-
gram was effective for community-residing elderly people.
From the Schools of ‘Occupation and Leisure Sciences and "Behavioral
Sciences, Faculty of Health Sciences, ‘School of Public Health, §Faculty
of Health Sciences, University of Sydney, Sydney, Australia; "Center for
Education and Research on Aging, Concord Hospital, Sydney, Australia; and
'Glebe Aged Care Assessment Team, Central Sydney Area Health Service,
This research was conducted with ﬁnancial support from the National Health
and Medical Research Council (Project Grant 107342), the Mercy Family
Center, Hornsby, and the Balmain Leagues Club, NSW, Australia.
Box 170, Lidcombe 1825, Australia.
Secondary analysis of subgroups showed that it was par-
ticularly effective for men (n = 80; RR = 0.32, 95 %
CI = 0.17-0.59).
CONCLUSION: The results of this study renew attention
to the idea that cognitive-behavioral learning in a small-
group environment can reduce falls. Stepping On offers a
successful fall-prevention option. J Am Geriatr Soc
Key words: accidental falls; elderly; prevention; cognitive-
behavioralg small-group intervention
Falls among the elderly represent a major economic and
social problem."3 Falls themselves and the belief that
one might fall in fall-risk situations can result in restriction
of mobility and activity, feelings of helplessness, loss of
confidence, depression, and institutionalization?
There is now good evidence that multifactorial inter-
ventions conducted by health professionals with skills in
geriatric medicine can prevent falls.‘ Multifaceted inter-
ventions have generally been consistent in showing an ef-
fect, particularly if they are targeting persons at risk5 and
include several intervention approaches.“9
Studies of behaviorally oriented educational interven-
tions using group processes have had limited success in
reducing falls.‘°*" One of the educational program trials“)
to reduce falls demonstrated a significant trend (16%
fewer reported falls), but overall, this trial was not suffi-
ciently effective to be clinically useful. There is support
that, for an important reduction in falls risk, an intervention
needs to demonstrate at least a 30% reduction.5*‘3 In
addition, the evidence base for falls prevention has only
been established over the previous 10 years, so the role of
educational programs in falls prevention warrants reexam-
In this paper, a study of the effectiveness of Stepping
On, ‘3 a small-group based educational program, is reported
lr\GS 522148"/‘-1494, 2004
Q. 2004 by the American Geriatrics Society
. . ....__...-___...;.-..=.:_'::_-. e’: -'_. ‘ -
(submitted for publication), included distribution of pro-
motional materials; health professional referrals; media ad-
vertisements and editorials in local newspapers;
Baseline assessment comprised a background question-
naire of demographics, fall and health history, a functional
measure of mobility and balance (the Get-up and Go
Test),‘5 and the Rhomberg test of balance with eyes open
and closed. Baseline assessments of secondary outcome
measures were also conducted: the 36-item Short Form (SF-
36) Health Survey, which measures a persons perception of
their health across mental and physical health domains;“
the Modiﬁed Falls-Efficlacy Scale"
tivities of daily
(MES), which assesses efficacy beliefs over a wider contin-
ural challenge than the MFES. The item “Walk down three
snowy steps without a handrail” was altered to “Walk down
three wet steps without a handrail.” A 10-point scale of
“not at all” to “completely confident" was used with the
MES and MFES. The falls self-efﬁcacy measures were of
interest as a secondary measure, being a core concept of the
program.” Also assessed at baseline and follow-up were
the Physical Activity Scale for the Elderly (PASE)‘9'2° and
the Worry scale,“ designed for use in community-dwelling
elderly to identify the degree to which aspects of daily lives
to enhance self-efficacy22
program. A team of content experts,
ment their content
options to risk man
ognizing that older ad
change. A va
ing theory to guide participants in exploring barriers
agement,23-2“ and the use of adult
to develop knowledge and skills, rec-
ults have the capacity for learning and
falls prevention to supple-
introduced key content
areas.” These included lower-limb balance and strength
exercises known to
with visual loss and
safety,3°'3‘ and co
be effective in fall prevention,“
shared and reinforced within the context of the group. Each
ramps) were practiced.
, averaging 12 partici
took place over a 7-week '
lasting 1.5 hours, occurred at the program venue.
Excluded (n = 57)
Declined to participate (n = 347)
Excluded because of
(n s: IH)
(n = 3 I0)
Stepping On sessions
Not visited (n -2 3 I) Lost to follow-up before None puended (n = 5)
One visit (n = 25) end or study (n = 25) 01: xession (n = 5'!
Two visits in = 94) 34 sessions tn = 6)
Died(n=7) Ssessionstns I7)
Withdrew (n = 6) 6 sessions (n t 45)
Lost contact (n = 5)
Nursing home (n = 6)
Cognitive decline (n = l)
7 sessions (n -= 79)
Did not complete ﬁnal
assessment (n = Z3)
(all full data collected)
research."3' The schedule consisted of a monthly tear-off
postcard calendar, with the subject's study identiﬁcation
number, and was preaddressed and stamped. Subjects were
asked to record an “N” on each day that they did not fall
and an “F” if they had a fall. If a fall was reported, the RA
ject to complete the schedule.
The RA, who was blind to group allocation, conducted
the 14-month follow-up assessment. The follow-up assess-
Loss to follow-up (Figure 1) for the primary outcome
(falls) was low, with only 8% '
falls calendar (P = .89).
The required sample size was based on the anticipated effect
of the intervention on fall rate, with a power of 80% and an
alpha of 5%. It was estimated that about 40% of control
1496 CLEMSON ET AL.
nd history of falls.
nown risk factors for falls.”
For secondary outcome variabl
PASE, and Worry scale) change sco
nus baseline score)
These are k
es (MES, MFES, SF-36,
res (follow-up score mi-
using independent sam-
—month period, clos-
nd six inner city met-
litan localities within Central Sydney Area Health
Service, two in eastern Sydney, and one in Newcastle, a
of Sydney. Fig '
reasons, 37% were ex-
uring a variety of functional
ble on the days the pro-
ks over the follow-up period, whereas the
status as measured using SF-36 scores (Table 2).
Median length of follow-up for all subjects was 429
days (range 2-529 days; interquartile range 418-429 days).
After 14 months of follow-up, there had been 255 falls in
the control group and 179 falls i
SEPTEMBER 2004—VOL. 52, NO. 9 pics
Up and Go test to define groups, sex, a ’ ' ' _
h . . ications and psychotropic drugs used and ‘ff
. . 0 . n t e lmervelzuon group’ recommendations arising from the follow-up home visit. At _' _ "
Eighty-nine (58 A») control subjects and 82 (52 A») program the end of 14 months 59% 1" = 77) of Program pam_Cl__
subjects reported one or more falls, and 53 (35%) control pants were still doing their exercise routinely although only " ii
. 0 .
sublects and 4.0 l26 /"l program sublects reported two or 41% (n = 53) were continuing to do the strength exercises
more falls during follow-up‘ There were three Control sub" with ankle cuff weights Agreater proportion of the subjects
__ _____ _______________, __ __ who did n ’
Table 1. Baseline Com ' ' '
parisons of Characteristics of Con-
trol and Program Subjects
period (program n=21 (72%)
, control n =13 (42%),
chi-square (x2) = 5.67, P =.O2). The total number of med-
C°”"°l P’°9’3m ications taken was similar at follow-up and baseline for
l” = 153) in = 157) both groups which a nonsignificant analysis of mean
Characteristic n (0/1 change scores conﬁrmed (P = 55), but control subjects
’ ° (n = 20,. 16 /o) were more likely to start taking a new psy-
Fema1e 113 (74) 117 174) Ch20tI'0plC drug than program participants (n =11, 8%)
Fans in 1135112 months (x f §.4, P= O4) Seventy percent (n = 80) of program
0 53 (35) 54 (35) participants adhered to at least 50% of the home-visit rec-
1 25 (16) 27 (17) omrnendations Recommendations included removing or
2 2 75 (49) 76 (43) modifying home fall hazards such as removing clutter, in-
History of stroke 27 (13) 27 (17) creasing lighting levels, applying nonslip tape to step edges
History of knee anhrms 52 (34) 57 (35) and mg pat ways The most commonly reported self-
History of hip fracture 15 (10) 9 (3) initiated action reported by the pa ’
Use psychotropic drugs 28 (18) 35
mobility (28%, n=153), such
Control (n = 153) Program (n == 157)
Mean :1: Standard
Age 78.47 1: 5.66 78.31 :t 5.26
Number of falls past 12 months 2.53 :1: 3.84 2.19 :l: 2.94
Number of people in household 1.50 i 0.81 1.61 i 1.04
Total number of medications 4.33 :1; 2.83 4.37 :1: 3.05
Days in bed previous 2 weeks‘ 0.19 i 0.73 0.43 :l: 1.68
Hospital admissions in past year 1.60 :l: 0.49 1.66 :l: 0.47
Short Portable Mental Status Questionnaire 9.77 :1; 0.58 9.78 i 0.50
Flhomberg balance. eyes closed‘ 1.76 i 0.82 1.74 :l: 0.82
Get up and go testi 2.11 :t1.11 1.92 :1: 0.99
Modified Falls Efficacy Scale 123.04 :1: 22.80 123.93 i 22.00
Mobility Efficacy Score 66.75 i 26.28 65.42 :l: 26.28
Physical Activity Scale for the Elderly score 79.19 :1: 54.01 75.52 i 43.93
Worry scale 0.43 i 0.45 0.47 :l: 0.55
SF-36 PCS‘ 38.79 :l: 10.74 38.37 :l: 10.84
SF-36 MCS’ 54.29 :l: 10.26 53.21 :1: 11.08
‘Missing data, max three persons.
:Get up and go measured as a rank—order‘scale.“
Analyzed using weights from the Australian Bureau of Statistics 1995 Health Survey.
SF-36 = 36-item short form; PCS = physical component survey; MCS = mental component survey.
scanning ahead for hazards when walking (other data not
reported). Although an at-risk population was targeted, it was a rel-
atively healthier, less-frail group than recruited in some of
the previous successful multifactorial interventions. For ex-
DISCUSSION 6 ,
The intervention group,s reduction in fans of 31% ample, one study excluded persons who did vigorous ac-
(P =.O25) indicates that the Stepping On program was ef-
fective for community-residing elderly people.
The analysis of RR for two or more falls, using more
traditional methods, supports the results for the negative
binomial regression, which accounts for multiple falls. The aged 80 and Older‘
This study places back on the agenda the viability and
efficacy of educational programs, at least a particular type
of education small-group program (one that uses a cogni-
tive-behavioral approach to increase knowledge and change
attitudes and behaviors). This program was based on an
active rather than a prescriptive approach, promoting per-
sonal control and problem solving, which provides oppor-
tunity for individuals to make behavioral changes.
The effectiveness of Stepping On in comparison with
Based on previous research,5'3’ recruitment targeted
people with a history of falling, and the resultshave sup-
ported this approach. The results add to the growing sup-
Subjects pe_r p
in study at end of
(152) (149) (147) (146) (139) (138)Control(n) The participants of the Stepping On program used
(156) (152) (152) (143) (147) (147) intervention (n)
Months from baseline assessment
Figure 2. Falls per month for control and intervention groups.
1492 CLEMSON ETAL. SEPTEMBER 2004—VOL. 52, NO. 9 mos
Relative Risk (95% Confidence Interval) .
Subject n 21 Falls‘ 32 Falls‘ All Falls’
ﬁll subjects 310 0.90 (0.73—1.10) 0.74 (0.52—1.04) 0.69 (0.50-0.96)
" Cumulative incidence ratio.
< 75 79 1.13 (0.76—1.69) 1.18 (0.59—2.34) 0.96 (0.50—1.85) ;
2 75 231 0.83 (0.66-1.05) 0.63 (0.42—0.94) 0.62 (0.43-0.89)
Female 230 1.03 (0.81—1.30) 0.83 (0.55-1.25) 0.96 (0.67—1.39) ; g,-
Male 80 0.61 (0.40-0.92) 0.56 (0.29—1.05) 0.32 (0.17—0.59)
Measure of functional mobility and balance .
Get up and go test score = 1 (normal) 132 1.94 (0.67—1.32) 1.06 (0.59—1.92) 0.76 (0.46—1.25) 1
Get up and go test score = 2, 3 148 0.83 (0.64—1.08) 0.50 (0.31—O.82) 0.56 (0.37—O.85) /$1
(slight to mild abnormal)
Get up and go test score = 4, 5 29 1.09 (0.54—2.22) 1.36 (0.54—3.42) 1.49 (0.44—4.99)
(moderate to severe)
History of falls in past 12 months _ 1
None 107 1.13 (0.76—1.69) 1.26 (0.51—3.14) 0.88 (0.50—1.54) gj
Falls 203 0.83 (0.66—1.05) 0.65 (0.46—O.94) 0.66 (0.46—0.95)
Negative binomial regression model.
activities. This was a particular focus of the Stepping On scale for the people recruited than those w
mobility or even housebound.‘3"”-42 Safely mobilizing
ut did not make a difference in the basic self-care type The difference in the contact time between the inter-
activities (MFES) It may be that those recruited were more vention group and the control group could have provided a
high y efﬁcacious, but comparison between baseline efﬁca- bias, ut because the adherence to the falls surveillance
cy scores and other published data”-2‘ do not suppo system was the same in both groups, bias from this cause
to a specific functional domain."° Unless the assessment most effect in its latter stages.
reviews what a particular program offers or what charac- Adherence results s
teristics are changed, it simply will not provide a result. It ponents of the program
may be more likely that the MES was a more appropriate home safety and the ex
. t of Stepping On Program on Self-Efficacy, Physical Activity, Protective Fall Behaviors, Health Status, and
“—"‘—"“‘—""““'—“*‘ Mean 95% Confidence Interval 1;?
Outcome Measure‘ n (Mean Change at Standard Deviation) Difference
of the Difference
Mobility efficacy scale 125 ( — 3.38 :l: 17.18) 133 (0.89 :l: 16.46) 4.28’ — 8.40 to — 0.54
Modified falls efficacy scale 125 (— 1.10 :l: 19.60) 133 (0.63 a 16.40) 1.74 — 6.14 — 2.67
FaB scale 126 (3.07 :l: 0.45)‘ 134 (3.19 :l: 0.35)‘ 0.121 — 0.21 to — 0.02
Physical activity scale 127 ( — 13.48 :l: 42.25) 132 (— 4.40 :l: 36.25) 9.08 — 18.70 — 0.54
SF-36 physical component 125 (0.68 :l: 9.04) 133 (— 0.02 :l: 8.34) 0.70 — 2.94 — 1.88
SF-36 mental health component 125 ( — 0.52 :l: 10.00) 133 (0.01 :t 9.65) 0.53 — 2.95 — 1.88
Worry scale 1 124 (— 0.01 :l: 0.27) 134 (— 0.05 :l: 0.73) 0.04 — 0.04 — 0.13
earning in a small-group environment
other successful fall-
to a nursing home. N Engl] Med 1997;337:1279-1284.
4. Gillespie LD, Gillespie W], Robertson MC et al.
than women. The earlier SAFE study
erly people living in the community N Engl J Med
10 also found a statis— ‘9943331=311‘327-
- - - - - . a , i , us L al rev n di abil' f I
tically significant effect (odds ratio = 0.53) for men aged 75 7 :rd:'°a'ﬂ’:s acid msomlid :“'a'l"5An: J '}f:'b’|:'cd Piﬁltﬂ
and older. The only secondary outcome 1994;84,3o0_3o5_
current study for which th ' ' ’ 8. DayL, Fildes E, Gordon 1 et
effect with sex was the PA
al. Randomised factorial trial of falls prevention
_ , among older people living in '
control group sh ' ' '
10. Hombrook MC, Stevens V], W" l '
munity-dwelling older persons: Results from a randomised trial. Gerontologisr
d women. One no- 19_94%34=“5‘23-
h less to be 11. Reinsch S, MacRa-e P, Lach
ty-eight percent (n = 38) of men and 19%
. tologist 1992;32:450-456.
_ . _ 12. Campbell A], Robertson MC, Gardner MM . '
en were living with a spouse (P< .001). The
trial in women 80 years and older. Age Ageing
support of a spouse may be conducive to the uptake and 1999i13=513'513v
follow through of fall-prevention strategies taught in the 13'
16. Ware JE
terpretation Guide. Boston, MA: Health
17. Hill KD, Schwartz JA, Kalogero
. Wisocki PA. Worry as a phenomenon relevant to the elderly. Behav Therap
, anada, May-June 1998.
. Field 1, Leiscester M. Lifelon ‘ '
don: Routledge Falmer, 2000.
. Campbell A], Robertson MC, Gardner MM et al. Randomised controlled trial
of a general practice programme of home based exercise to prevent falls in
elderly women. BM] 1997;315:1065-1069.
. Ivers RQ, Cumming RG, Mitchell P mpairment and falls in older
adults: The Blue M ‘ .
. Cumming RG. Epidemiology of med‘
elderly. Epidemiology l998;12:43-53.
. Cumming RG, Klineberg ' ' ‘ 'cs and hip frac-
tures in the elderly. Med J Aust 1993;158:414-417.
34. ' , . development of an assessment to 44. Devor M, Wang A, Renvall M et al. Compliance with social and safety rec.
evaluate behavioral factors associated with falling. Am] Occup Tl1erap 2003; ommendations in an outpatient comprehensive geriatric assessment program_
57:380-388. J Gerontol l994;49:M168—M173.
35. Glynn R], Buring]E. Ways of measuring rates of recurrent events. BM] 1996; 45. Fabaeher D, Josephson K, Pietrus
36. Robertson MC, Gardner MM, Devlin N et al. Effectiveness and economic
37. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among elderly per-
sons living in the community. N Engl] Med 1988;319:1701-1707.
38. Mendes de Leon C, Seeman T, Baker D et al. Self-efficacy. physical decline and
change in functioning in community-living elderly: A prospective study.
J Gerontol B Psychol Sci Soc Sci 1996;513:5183-5190.
39. Cumming RC, Salkeld G '
Reducing Falls and Building Conﬁdence:
Session 1: Introduction, Overview, and Risk Appraisal
Building trust, overview ‘ '
Generate strategies to get around in the local community and reduce the risk of falling. Learn about the features of a safe
shoe and identify clothing hazards.
Session 7: Review and Plan Ahead
Express personal accomplishments from the past 7 weeks and reflect on the scope of things learned. Review anything
requested. Finish any segment not adequately completed. Determine safety strategies to protect against bag snatching.
' ' ure.
Identify strategies to assist in safely using trains. Time for farewells and clos